Ghost Peak- UPLC

[kalsang]

Recently I have successfully transferred HPLC method to UPLC through geometrical scaling. However, I’m stuck with an appearance of ghost peak at the RT (3.4 min) of the main peak. I have been struggling to solve this problem from past many days but failed.

Some my studies includes- washing column with 100% organic, changing water, priming injectors and all lines, and gave zero injection (ghost peak still appears!).

Following are my chromatographic conditions-

Buffer= 10 mM K2HPO4; pH adjusted to 7.6 with orthophosphoric acid
Mobile phase A = Buffer:CH3CN (90:10)
Mobile phase B = CH3OH: CH3CN (40:60)
Column = BEH-Shield RP-18 (100x2.1)mm 1.7 µm
Diluent = Buffer: CH3OH (80:120)
Column temperature = 40 Degree celcius
Wavelength = 210 nm
Run time = 9.52 min

Gradient profile: (Time:%B) (0:5) (0.41:10) (1.36:20) (2.72:35) ( 7.48:55) (8.16:5) (9.52:5)

Please give your company to solve this problem. I must get rid of this ghost peak in order to validate the method. I can’t change gradient profile as it is a transferred method from HPLC.

[rwjohnson]

I'm certainly no expert on troubleshooting, but it sounds like carryover from the injector somehow. Do you have a rinse on you UPLC?

What type of UPLC? I have been using an Agilent 1290 and the automatic rinse port seems to work really good, except that now I am using a different CDS and it doesn't allow for advanced controls of the autosampler and it has to use a rinse vial. I have also experienced carryover on similar-style HPLC units where the main analyte will contaminate the needle and other places. I have also experienced contamination build-up on the underside of my rinse vial.

Good luck.

Ron J.

[kalsang]

Hi Ron,

Thanks for the reply.

I am using ACQUITY UPLC® System which have both weak and strong needle wash. I think injection carry over is almost rule out because, I have primed the injector 6 cycles and moreover, my weak (50:50 water:CH3CN) and strong (80:20 water:CH3CN) needle wash will remove any carryover.

Now I didn't understand when you said the contamination in other places?

[rwjohnson]

Regarding the other places, I found that material would get carried from a sample vial by the needle and had some success by cleaning all around the needle assembly, guide and seat. These were items that wouldn't typically get cleaned through the rinses and cleaning. The contamination happened as the needle was removed from the sample vial and before injection.

I think carry-over would be a likely culprit if the ghost peak has same RT as primary analyte. Is it really a significant size compared to your sample and standard?

Ron J

[jdelosrios]

Dear kaljang,

I suppose you have checked extensively the carryover possibility. I have had similar problems with Acquity UPLC's and they are really annoying and not easy to detect the source. If you haven't tried yet, you can change the injection mode (full loop, partial loop, or partial loop with needle overflow),or even change the loop, because sometimes this piece retains part of the analyte in the connectors that is injected every time in the system.

Good luck

[kalsang]

I have checked the carry over many times. Even if I give zero injection continuously, I get that ghost peak. Will the carryover from the loop still appear after giving zero injections for many times?

[tom jupille]

Try running the "three blank gradient" test as described here:
http://www.lcresources.com/resources/TSWiz/hs400.htm

If the ghost peak increases in area in proportion to the equilibration time, then the source is contaminated "A" solvent.

[Alexandre]

I agree. If you not injecting anything and just profiling gradient and see the peak, then the problem is in the mob phases.

However if the system is already contaminated it may be from your previous injections. If peak get smaller and smaller with no injections, then it is from the system. If stays the same – not conclusive, it may be from the mobph and/or from carryover. You may need hundreds injections until you see the drop if it is from a deposit somewhere inside.

Important to remember, your new peak may be always was there, hidden in HPLC under your old peak. With better resolution it will appear in UPLC as a distortion. It may precipitate and drags though the all your washes. In this case you should have carryover effects in the past with HPLC. Again, you translated geometrically, but the stationary phase is not exactly the same, is not it?

The number of washes (6 washes) does not automatically guarantee that you wash well. Your strong wash should be stronger than sample solvent, not the UPLC phase. Some stuff is sticky, we use base as a strong wash to get rid of persistent tolfenamic acid for example. Even if your sample was in strong organic solvent X the strong organic solvent Y may be not compatible, you need to match the chemistry exactly.

I am sure you filter samples and change mob phase as recommended. Some observation, your sample is in 60% MeOH and your staring grad. condition is something like 80% water – this is the recipe for disaster. When these 2 mix together you may have a precipitation. For critical applications the staring solvent grad % org. should be the same as your sample diluent.

Keep us updated, we all learning.

[kalsang]

It seems my problem is almost solved but I am sorry to say that I could not trace the exact source of problem because I didn't followed systematic approach in tackling it. Just like any other day, I came to the lab and prepared a fresh mobile phase A as usual and saw in the subsequent run that the ghost peak has decreased drastically. Although earlier I used to get ghost peak with the change in mobile phase A.

I must blame the buffer salt or pH meter probe or quality of the water, is obscure to me. Now what I fear is the re-appearance of the ghost peak if I prepare a mobile phase A another time.

@Alexandre, the column chemistry is not same but equivalent (in fact the most equivalent) since I referred Waters column selectivity chart. My sample is soluble in water, so it wont precipitate in contact with initial high quantity of water?? I think I must keep 100% water as a strong wash??